YouTube has announced that it will begin automatically detecting and labelling videos that contain significant photorealistic AI-generated content. The platform will now use internal signals to identify such content rather than relying solely on creators to disclose the use of artificial intelligence themselves.
This move reverses a previous policy that depended on voluntary creator disclosure, a system that had been in place since 2024. As part of this update, YouTube is also making AI labels more prominent.
Despite these changes, certain types of content may still hide their origins. Videos that are animated, unrealistic, or contain only a small amount of AI-generated material may not be automatically labelled.